|Publication number||US8024404 B2|
|Application number||US 10/630,472|
|Publication date||Sep 20, 2011|
|Filing date||Jul 30, 2003|
|Priority date||Jul 19, 1999|
|Also published as||CA2380141A1, CA2380141C, DE60042662D1, EP1456772A2, EP1456772B1, EP2105849A2, EP2105849A3, US6859821, US7490126, US20040024820, US20050102364, WO2001006364A2, WO2001006364A3, WO2001006364A8|
|Publication number||10630472, 630472, US 8024404 B2, US 8024404B2, US-B2-8024404, US8024404 B2, US8024404B2|
|Inventors||Raymond E. Ozzie, Jack E. Ozzie|
|Original Assignee||Microsoft Corporation|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (52), Non-Patent Citations (14), Referenced by (4), Classifications (28), Legal Events (3)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application is a division of U.S. patent application Ser. No. 09/356,148 filed Jul. 19, 1999. The following commonly-assigned U.S. patent applications are related to the present application, and incorporated by reference herein in their entirety:
The invention relates generally to peer-to-peer collaboration systems, and more particularly, to a method and apparatus for assigning designations to combinations of collaborating computers and user operating those computers in order to facilitate establishing and maintaining data consistency among collaborators in such systems.
The Internet has established a dynamic, public environment for communication and interaction among its millions of users. In business, the Internet has redefined vendor-manufacturer, manufacturer-distributor, distributor-customer, and other relationships. With extension of the Internet technology into internal, secured networks of individual companies, the “intranet” or “private Internet”, as it is called, has enabled new forms of document and information sharing between individual employees and work groups using company directory and network infrastructure. On-line services, such as electronic bulletin boards and chat rooms, electronic commerce, and technical support for products, are available on the World Wide Web (“WWW” or “Web”) operating over the Internet.
The Internet has, at its core, a server-client architecture, in which individual clients (i.e., Internet-content users) interface via computer-executable applications such as browsers with servers (i.e., Internet-content providers) to obtain documents from Web sites. Browsers are software programs that enable personal computers to request, receive (e.g., download), interpret, and present Internet documents, and generally navigate the Internet. Web servers typically have standard interfaces for running external programs, the most common is the Common Gateway Interface (CGI). Web sites are collections of documents, usually consisting of a home page and related, linked documents, located on servers remote from the client. The documents can be compound documents, containing data, graphics, video, sound, and/or other types of media, as well as links to other documents. Essentially, the WWW is a web of interconnected documents, or, more precisely, document objects that are located at various sites on the Internet.
Among the types of document objects on the WWW are documents and scripts. A script is an executable program, or a set of commands stored in a file, that can be run by a Web server to produce a document that is then returned to the Web browser. Typical script actions include running library routines or other applications to get information from a file or database, or initiating a request to get information from another computer, or retrieving a document corresponding to a selected hypertext link. A script is run on the Web server when, for example, a user selects a particular hypertext link in the Web browser.
Underlying the Internet technology are advances in standardization, including personal computer hardware, software, network protocols, and infrastructural conventions (such as the “Uniform Resource Locator” or “URL”). URLs provide location addresses for all document objects on the WWW. A URL uniquely references a document object and often defines an access algorithm using Internet protocols.
“Internet protocols”, as that expression is used herein, are defined as current or future language conventions and present or future communication protocols that are generally-accepted and used for generating messages including documents for transmission over the Internet, or for transmitting such messages over the Internet, respectively. Such language conventions include, for example, at present, Hypertext Markup Language (“HTML”) and eXtensible Markup Language (“XML”). Such communication protocols include, for example, at present, Hypertext Transfer Protocol (“HTTP”), TCP/IP, FTP and GOPHER. Those skilled in the art will be familiar with these language conventions and protocols.
HTML is a language used for writing hypertext documents, containing multimedia content and links to other documents. An HTML document includes a hierarchical set of markup elements, where most elements have a start tag, followed by content, followed by an end tag. Tags are enclosed in angle brackets (“<” and “>”) and indicate how the document is structured and how to display the document, as well as destinations and labels for hypertext links. There are tags for markup elements such as titles, headers, text attributes such as bold and italic, lists, paragraph boundaries, external links to other documents, and internal links to other parts of the same document, graphic images, and many other document features. Many books are in wide circulation regarding programming using HTML.
XML is another of the language conventions included in the Internet protocols. XML is compatible with and complementary to HTML, at least in its current version. It is a standard way of describing a class of data objects as stored in computers, called XML documents, and behavior of programs that process these objects. XML documents are made up of storage units called entities, which contain either text or binary data. Text is made up of characters, some of which form the character content of the documents, and some of which form markup. Markup encodes a description of the document, such as its storage layout and logical structure. A software module called an XML processor is used to read XML documents and provide access to their content and structure. Further information regarding XML can be had with reference to Version 1.0 of the XML specification, available at <HTTP://www.w3.org/XML>, and incorporated herein by reference.
A Web server and a Web browser communicate typically using the HTTP message protocol and the underlying TCP/IP data transport protocol of the Internet. In HTTP, the Web browser establishes a connection to a Web server and sends an HTTP request message to the server. In response to the request message, the Web server checks authorization, performs any requested action such as downloading a document identified by its URL, and returns an HTTP response message containing either the HTML document resulting from the requested action or an error message. The returned document may simply be a static file stored on the Web server, or it may be generated dynamically using a script called in response to the request message.
To take advantage of the Internet, tools and resources have been developed in compliance with the Internet protocols, including company-critical applications such as e-mail. E-mail is electronic mail, by means of which documents are sent and received electronically at selected addresses. It has been estimated that a vast majority of Internet-based interaction is through the use of e-mail and other browser-based media that follow a “document send and receive” model. Perhaps due to that model, users often view the Internet as inherently “peer-to-peer”, with individuals accessing documents provided by other individuals, without intervention by a higher authority.
The Internet is dynamic and flexible in providing users with entertaining and useful ways of communicating, though it does not meet all the needs of users. While users interact increasingly through the Internet, they continue to interact “off” of the Internet in more conventional ways, such as through multi-medium (phone, fax, whiteboard), multi-temporal (real-time, overnight mail) and other informal means of communication.
It would be desirable to extend the Internet paradigm to personal and private communications and other shared and mutual activities between individuals and small groups in shared private spaces. Such interactions should preferably occur instantly, directly, and confidentially between participants' personal computers, or other network-capable devices, without exposing such interactions to server bottlenecks through which Internet traffic is typically funneled to third-party Web sites where communications can be intercepted and confidences violated.
It would also be desirable to provide a technique that allows users at various remote sites to share and edit documents on a peer-to-peer basis, while maintaining consistent copies of the documents at local sites. Such a technique should permit any of the users to issue change requests regarding the documents, which changes can then be made asynchronously to the copies at all the local sites in a coordinated fashion even if the sites are disconnected from the network when the change requests are issued. The technique should provide a high level of data consistency across all the sites so that each site presents essentially the same document for viewing and/or editing by the various, geographically dispersed users.
In accordance with the principles of the invention, a data change request priority scheme uses request sequence number information and dependency information to make data changes. In case of a dependency collision between data change requests, the ordering of execution of deltas in the event of a collision depends on the endpoint (a unique pairing of a device and a person) that originated the delta. In particular, each endpoint that is a member of a telespace is assigned a unique endpoint number, corresponding, for example, to the order in which the endpoint originally joined the telespace.
In one embodiment, a unique number is assigned to each endpoint of each of the telespace members. Each endpoint number indicates the order in which the member joined the telespace, and, for each member invited to join (or otherwise sponsored) by another telespace member (the “inviting member”), it also indicates the inviting member. As more and more members are invited to join by those new members, the designations add additional orders of digits to indicate the family tree or chain of inviting members. Accordingly, the designations can be thought of as indicating the ancestry of each member added to a telespace after the founding members of a telespace. Thus, where the designations comprise a number of orders, the first order designates a founding member of the telespace, and each successive order after the first order (e.g., second, third, etc, order) designates a member invited to join the telespace by an inviting member specified by the preceding order digit. The inviting member may be a founding member or another member of the telespace of intermediate seniority and with telespace membership sponsoring privileges. In one embodiment, each inviting member is responsible for assigning telespace numbers to members it sponsors.
In another embodiment, the endpoint numbers are used to resolve dependency collisions between data change requests.
The above and further advantages of the invention may be better understood by referring to the following description in conjunction with the accompanying drawings in which:
It may prove useful to introduce at the outset certain terminology and abbreviations used in this specification, including: activity, client, communications manager, component, controller, delta, device, dynamics manager, engine, framework, identity, member, navigation, person, server, telespace, tool, URL, and XML.
B. Conventional Computer System
The computer system 100 includes a central processing unit (CPU) 105, which may include a conventional microprocessor, random access memory (RAM) 110 for temporary storage of information, and read only memory (ROM) 115 for permanent storage of information. A memory controller 120 is provided for controlling system RAM 110. A bus controller 125 is provided for controlling bus 130, and an interrupt controller 135 is used for receiving and processing various interrupt signals from the other system components.
Mass storage may be provided by diskette 142, CD-ROM 147, or hard disk 152. Data and software may be exchanged with client computer 100 via removable media, such as diskette 142 and CD-ROM 147. Diskette 142 is insertable into diskette drive 141, which is connected to bus 130 by controller 140. Similarly, CD-ROM 147 is insertable into CD-ROM drive 146, which is connected to bus 130 by controller 145. Finally, the hard disk 152 is part of a fixed disk drive 151, which is connected to bus 130 by controller 150.
User input to the computer system 100 may be provided by a number of devices. For example, a keyboard 156 and a mouse 157 may be connected to bus 130 by keyboard and mouse controller 155. An audio transducer 196, which may act as both a microphone and a speaker, is connected to bus 130 by audio controller 197. It should be obvious to those reasonably skilled in the art that other input devices, such as a pen and/or tablet and a microphone for voice input, may be connected to client computer 100 through bus 130 and an appropriate controller. DMA controller 160 is provided for performing direct memory access to system RAM 110. A visual display is generated by a video controller 165, which controls video display 170.
Computer system 100 also includes a network adapter 190 that allows the client computer 100 to be interconnected to a network 195 via a bus 191. The network 195, which may be a local area network (LAN), a wide area network (WAN), or the Internet, may utilize general-purpose communication lines that interconnect multiple network devices.
Computer system 100 generally is controlled and coordinated by operating system software. Among other computer system control functions, the operating system controls allocation of system resources and performs tasks such as process scheduling, memory management, networking and I/O services.
C. Conventional Approach to Controlling a Data Model
The conventional approach for controlling a data model as exemplified in the system 200 was appropriate in an age in which computer systems, whether mainframes or personal computers, operated in relative isolation. With the advent of the Internet, computer users demand improved ways of collaborating, and the invention provides what can be called “model mediated control”, in which a dynamics manager mediates changes effected by the data-change engine to the data model.
D. Internet Paradigm for Intercomputer Operation
As will be explained in more detail below, such peer-to-peer communications can be made directly or via a relay device 316. The relay device 316 is preferably a “store and forward”, which can store messages destined to a peer unit 314 which is temporarily disconnected from the Internet, and later, on reconnection, can forward the messages to that peer unit.
E. ABC system Architecture and Operation
The framework 402 includes a user interface manager 408, identity manager 410, telespace manager 412, activity manager 414, storage manager 416, controller services manager 418, dynamics manager 420, and communications manager 422.
The user interface (UI) manager 408 is responsible for managing shared services for a number of user interface controllers (not separately shown). The UI manager 408 manages layout of activities within panes of a display window, and otherwise provides a desired “look and feel” for the user interface. The UI manager 408 also manages activity navigation (for example, go to, next, previous, etc.) and maintains a navigation history.
The identity manager 410 is responsible for maintaining identities of a telespace's member. As noted above, an identity is the name, and corresponding URL, by which each user is known by others. Individual users may have one or many identities. The identity manager 410 maintains a record or table, preferably in XML, of the identities. The identity manager 410 can also maintain a record or table, preferable in XML, of the URL's of telespace members and their corresponding device URL's. Alternatively, a separate member manager can be implemented.
The telespace manager 412 is responsible for managing each of the telespaces 404 that may be opened on the peer unit 314A. Each telespace 404 is an instantiation of one or more activities. Each telespace 404 has a corresponding activity manager 414.
The activity manager 414 is responsible for (a) adding new activities to a telespace, (b) opening existing activities in a telespace, and (c) updating telespace activities from new activity template versions. To add a new activity, the activity manager 414 is provided with the URL of an activity template, opens the template, and extracts and propagates the template information (such as component URLs) into the telespace. The template defines the initial activity configuration for a telespace. A user may add additional activities to the telespace 404 later as needed. After being added, an activity is “part of” the telespace and visible to all telespace members. A telespace has a tag to identify its activity manager and bind the activity manager and data. Preferably, each document has a local registry linked to it, with XML tag names maintained in the registry to express mapping (reference pointers or associations) in an extensible, platform-independent way, between the document and its corresponding telespace. Each telespace member has a framework and an activity template for the telespace available on his or her peer unit 314A-D.
Each activity includes a tool, such as tool 424 and an engine, such as engine 426. The tool 424 provides the user interface (UI) capability for the activity, interacting via UI 406 with a member. UI interaction may include UI events initiated, for example, via the keyboard 156 (
For creation of an activity template, a software developer may write or adapt a tool and engine for use within the framework. An activity template is a persistent representation of the tool and engine components comprising an activity. An activity template can be distributed, for example, as shrink wrapped software or downloaded, for example, over the Internet to peer unit 314A from a remote server. Activity components can be regarded as Web documents and are represented persistently via URLs. The activity template itself preferably has a URL, which allows for tracking activity design changes. The activity template can be a single activity template or an activity collection template. A single activity template pertains to only one activity, such as “chat”. An activity collection template pertains to a collection of activities, such as “chat and outline”.
For use, the ABC system 400 gets a member's identity via the identity manager 410, opens a telespace manager, requests the telespace manager to open a telespace via a URL, requests the telespace manager for an activity manager, and, then, the activity manager opens an activity, typically by using the activity's URL. Then, the ABC system 400 is ready for members to use the telespace to perform the shared, focused tasks offered by the particular activity.
As shown for peer unit 314A, the ABC system 500 includes a number of illustrative telespaces in which peer unit 314A is subscribed as a member, including a car designing telespace 502, a chat telespace 504, and a chess playing telespace 506. As also shown, peer unit 314 has a number of components used during operation of the telespaces 502, 504, 506. Peer unit 314A has a memory 510 that maintains a telespace record 508, which lists and describes each of the telespaces 502, 504, 506 in which it is member, and a member record 512, which lists and describes each of the identities of telespace members that may participate in those telespaces. A UI 512 can receive user-initiated changes to the telespace and member records 508, 514.
Each of the telespaces 502, 504, 506 instantiates a different one of the illustrated activities 522, 524, 526. (Alternatively, the illustrated activites 522, 524, 526 could reside in a single telespace, in which case all members of the single telespace would have access to each of the activities.) Each activity 522, 524, 526 includes a respective tool 532, 534, 536 specific to the activity, and under control of a respective engine 542, 544, 546 for carrying out changes to a corresponding data model 552, 554, 556 contained in memory 510, and which is persisted in each of the telespaces 502, 504, 506. For example, the car design telespace 502 is an instance of an activity 522 that can include a tool 532 in the form of CAD (computer-aided design) software, and the chess-playing telespace 506 is an instance of an activity 526 that can include a tool 536 in the form of a software chess game.
Users pursue the activities 522, 524, 526 of the telespaces 502, 504, 506, via the user interface 514, which interfaces with the tools 532, 534, 536 for providing the tools with user requests. Responsive to the requests, the engines 542, 544, 546 change the persistent state of a corresponding data model 552, 554, 556 in memory 510, under the control of the dynamics manager 501 to reflect a current state. For example, the persistent state of the data for the chess-playing telespace 506 might include a chess board and locations of the chess pieces on the board; a user request may specify a chess piece move, which constitutes a change in the data, i.e., a delta; and the current state would reflect the execution of the delta and the consummation of the move.
More specifically, as indicated by line “a′”, the tool 536 initiates creation of a delta by requesting its generation by the dynamics manager 501. The dynamics manager 501 creates the delta and returns it over line a′ to the tool 536, and may record the created delta in case fault recovery is required.
As indicated at line “a” of
As indicated at line “b” of
The dynamics manager 501 includes queue structures 612 and a dynamo process 614, preferably a computer-executable program containing the main logic of the dynamics manager. The dynamics manager 501, and its dynamo process 614, are in bi-directional communication with the communications manager 622. The dynamics manager 501 enqueues the received deltas in the queue structures 612 for ordering purposes, then the dynamo process 614 services the queue structure in processing the enqueued deltas, and directing the engine 546 to execute them in an appropriate order. More specifically, the dynamo process 614 of dynamics manager 501 sends a control signal in the form of a “delta command execution” message through the interface 615 to the dynamics manager end 616 of the engine 546 over line “c” of
The dynamo process 614 of dynamics manager 501 also disseminates all locally-initiated deltas to other endpoints via the communications manager 622, which transmits them over the network, e.g., 626 to the peer units 314B-D. At each peer unit 314B-D, the respective dynamics manager enqueues received deltas in its local queue structures, and then passes them in a specified order to its respective engine for execution.
A single dynamics manager can support a single telespace or multiple telespaces, depending on the application, and thus can be responsible for directing a number of engines to execute deltas for the various telespaces. In systems having multiple telespaces, as illustrated in
Finally, as indicated at line “d” of
Preferably, at substantially the same time that the change is being made by engine 546 to the data model 556 stored in memory 556 of peer unit 314A, the other peer units 314B-D are making the analogous change to their locally stored data so as to effect the intent of the delta in a consistent manner. Situations may arise where one or another of the peer units 314A-D is not connected to the network, e.g., the Internet. Where the peer unit 314A may be disconnected, the communications manager 622 employs a device presence detector 430 to ascertain whether the peer unit 314A is connected and, if not, stores outbound deltas in an outbound delta store 628 until such time as connection is restored, and then transmits the stored deltas. Where one or more destination peer units 314B-D are disconnected from the network, the communications manager 622 of peer unit 314A can proceed to transmit the outbound delta to relay 316 (
It can be seen that, as between the tool 536 and the engine 546, user interface activities are effected directly only via the tool 536 and data model changes are effected directly only via the engine 546. In the vernacular, the tool 536 “owns” the UI and the engine 546 “owns” the data model. This can be contrasted with current, typical applications as illustrated in
F. Communications manager, Presence Server, and Relay
As noted, each communications manager 806A-D is responsible for managing all communications between its peer unit and the other peer units. For example, communications manager 822 manages bi-directional communications between peer unit 802A and peer units 802B-D. The communications can include messages sent either from the peer unit 802A to one or more of peer units 802B-D, or from one of the peer units 802B-D to peer unit 802A. The transmissions can be made directly and without intermediaries, provided the destination peer unit(s) 802B-D (in the former case) or peer unit 802A (in the latter case) are connected to the network, e.g., the Internet (“on-line”). As noted above, a device presence server 812 can be employed to ascertain whether the intended destination is on-line, or temporarily disconnected (“off-line”). If the destination is off-line, the communications can be made via a relay 814 (as noted above), which then forwards the message on to its destination when it returns on-line.
Thus, if peer unit 802A seeks to send a message to peer unit 802C, and the presence server 812 has informed the communications manager 806A that peer unit 802C is on-line, then the communications manager 806A sends the message via the network, e.g., the Internet 803, directly to the URL of peer unit 802C. At peer unit 802C, the message is received by communications manager 806C, which passes the message to dynamics manager 804C.
On the other hand, if peer unit 802A seeks to send a message to peer unit 802C, and the device presence server 812 has informed the communications manager 806A that peer unit 802C is off-line, then the communications manager 806A sends the message via the network, e.g., Internet 803, to the URL of the relay 814. The relay 814 can be implemented as a high-performance file server, which stores the en route message until peer unit 802C returns on-line, and then forwards the message to the URL of peer unit 802C. At peer unit 802C, the message is received as noted before, and passed through communications manager 806C to dynamics manager 804C.
Accordingly, the presence server 812 can serve the purpose within the ABC system 800 of monitoring whether the peer units 802A-D are on-line, and notifying each of the peer units of the connection status of the other peer units. To fulfill this purpose, the presence server 812 can be configured, for example, as shown for computer system 100 of
During operation, each of the peer units 802A-D can be responsible for notifying the presence server 812 of its connection status, i.e., whether it is on-line or about to go off-line. Alternatively, the presence server 812 can poll the peer units 802A-D from time to time to obtain their connection status. This information is stored in the presence server 812 in a preferably volatile memory 816 in a connection directory 818. The connection directory 818 is initially empty, and remains so until the peer units 802A D provide the presence server 812 with their connection status, e.g., provide connection status notifications. Then, upon receipt of the status information, the connection directory 818 stores the notifying peer unit's URL and its connection status.
The presence server 812 can also provide a connection-status subscription service for the peer units 802A-D. The presence server 812 will notify each subscribing peer unit 802A-D of the connection status of each other peer unit 802A-D, and of a change in that connection status. The presence server 812 can send the connection status notifications to the subscribing peer units 802A-D either (a) upon request of status change notifications from the peer unit 802A-D “wishing” to send a message; (b) upon connection of the peer unit to the network, e.g., the Internet 803; (c) from time to time, e.g., upon a change in status of any of the peer units; or (d) a combination of the foregoing. Preferably, all communications are made pursuant to Internet protocols. These protocols can include, among others, those mentioned above.
Each peer unit 802A-D can also subscribe to a membership-status service, and, by so doing, thereafter learn on an on-going basis of the membership of each telespace in which it is a member. To enable this service, the presence server 812 has a member directory 822 maintained in memory 816 for storing records for each member supported by the presence server. Each peer unit provides the presence server with its current membership status from time to time or, alternatively, upon any change it its membership status.
To illustrate, if peer unit 802A wishes to send a message, such as a delta, to peer unit 802B, peer unit 802A accesses the presence server 430 and obtains the URL of peer unit 802B to obtain its connection status. Assuming peer unit 802A subscribes to the subscription service, it will have received notification of the status of peer unit 802B. Thus the dynamics manager 804A will have passed the message to the communications manager 806A without knowledge of whether the destination peer unit 802B is connected to the network.
G. Distributed Data Consistency Model and Delta Processing
As noted above, the dynamics managers 804A-D in the peer units 802A-D are responsible for receiving locally-generated and remotely-generated deltas and maintaining a proper order for their execution by their respective engines 807A-D so as to maintain a high level of consistency of the data models across all the peer units 314A-D of the ABC system 800 participating in a telespace. This is made all the more difficult because deltas originating from one or another of the peer units may be received and/or executed out of order, or even after significant delays due to peer units being temporarily disconnected from the network. Moreover, the deltas may be generated based on different and inconsistent local data states.
Theoretically, data consistency can be described in terms of three properties, each achievable to some degree in collaboration systems, namely causality preservation, intention preservation, and convergence. More specifically, these properties of consistency can be described as follows:
Causality preservation (i.e., sequential consistency) requires that, if delta
A causes delta B, then delta A must be executed first so that the sequence that gave rise to the creation of the deltas is preserved in their execution. In other words, satisfaction of this property requires that the execution order of dependent operations be the same as their natural cause/effect order.
Intention preservation (i.e., semantic consistency) requires that the meaning of the deltas and the intention behind their creation be preserved. In other words, satisfaction of this property requires that, for any pair of deltas, e.g., delta A and delta B, the execution of each must preserve the intention of the other. To achieve this, the system records the data state at the time a user requested a change to that data state. For example, in a shared document, when change “x” was requested, the next word was “y” and the same is true when “x” is executed. Put another way, the effects of executing a delta on the data stored at each peer unit must be the same as the effects of executing the same delta on the data stored at the originating peer unit.
Convergence requires that the copies of data maintained by each and every peer unit 802A-D must be identical after executing the same deltas. This is known as “total ordering.”
The ABC system 800 records the necessary information to maintain a high level of data consistency and imposes delta execution order constraints on deltas whose execution is dependent on the prior execution of other deltas. For example, where deltas are designated with capital letters A, B, C, D, assigned in the order of delta receipt by the dynamics manager 804A of peer unit 802A, it is possible that the deltas A, B, C, D may be properly executed in that order of receipt to preserve causality. This may even be the case if, e.g., deltas A and C originated locally from peer unit 802A and deltas B and D came from remote peer unit 802D. In some cases, however, execution of those deltas in the order of receipt will result in the changes being made out of order giving rise to a causality violation. This may arise, for example, because of differential latencies in the communication paths traversed by locally-generated and remotely-generated deltas, which can result in the deltas being received in an interleaved fashion. Moreover, even if causality preservation is achieved, the result may violate the intention preservation property, and the changes may even be nonsensical. For example, delta B may seek to insert “four” before “legs” in the sentence “A cow has legs.” If delta A sought, without knowledge of delta B's change, to change “legs” to “a head” in that sentence, and is executed before delta B so as to change the context for delta B, the result is “A cow has four a head.” Indeed, this result makes no sense, and nor would the result if delta A were processed after delta B. As for convergence, a number of conditions may compromise the “sameness” (i.e., replication) of the copies of the data at the peer units 802A-D. For example, communication faults in the network may prevent deltas from being received by some peer units 802A-D, causing different delta execution results. It is difficult, if not impossible, to achieve total convergence, causality preservation, and intention preservation in a practical system, but a high level of data consistency is desirable for collaboration between the peer units.
In order to achieve a high level of data consistency, each dynamics manager 804A-D preferably creates deltas of a predetermined format, and records in the deltas and in their respective memories certain information used to maintain data consistency. Specifically, each dynamics manager 804A-D assigns to each delta it creates a sequence number (“Seq.No.”) to indicate the sequencial position of the delta among all deltas created by that dynamics manager at the particular peer unit 802A-D. The dynamics manager 804A-D then records the assigned sequence number in the delta along with the endpoint identifier (“EP”) specifying the endpoint (peer unit and person) that originated the delta. The dynamics manager 804A-D also records in the delta at the time of its creation information specifying a single dependency. The dependency information recorded in the delta is, e.g., in the form of the sequence number and EP of the endpoint originating one other delta, for example, usually, the delta last executed immediately prior to the creation of the subject delta by the engine 807A-D for the same activity as the subject delta and at the originating peer unit for that delta.
Some deltas do not depend on prior deltas, for example, because the state of the telespace data is unimportant to their execution. They are called “independent deltas”. For independent deltas, no such dependent sequence number need be provided. Each dynamics manager 804A-D extracts consistency information from the headers of received deltas, and maintains the information along with the rest of the delta information in memory.
The header 902 specifies consistency information, for example, as follows:
The vector 1006 is important in this regard for storing information used in maintaining a high level of data consistency, and thus deserves further discussion. The vector 1006 stores, for example, the following consistency information regarding each received delta: (a) an originating endpoint identifier (“UID”), which is an identification of that endpoint that is unique across the telespace; (b) an endpoint-relative sequence number, which is a running count of deltas in the telespace received from each end-point, and corresponds essentially to a per-endpoint time stamp for ordering purposes of the received deltas; (c) a dependent sequence number for the deltas, which indicates, as of the time an endpoint created a delta, it did so based on a certain prior delta in the sequence that had been executed by that peer unit (i.e., the “user seen” current state). In the preferred embodiment, all of this information can be extracted from the received deltas.
During operation, when a delta is received, for example, by the dynamics manager 804A, the dynamics manager places the received delta in either one of the local delta queue 1012 or the inbound delta queue 1014, depending on whether the received delta was locally initiated or remotely initiated. Some implementations can store only a pointer to each such delta in the appropriate queue. The queues 1012, 1014 are implemented, for example, as FIFO (first-in-first-out) workload buffers. (Alternatively, a single queue can be utilized in some implementations for holding both inbound and local deltas.) The dynamics manager 804A processes the deltas from the queues 1012, 1014 on a first in, first out basis. Some implementations can use an appropriate priority algorithm for processing deltas from one or the other of the local or inbound delta queues 1012, 1014, for example, for load balancing and/or latency control. It may be desirable to process all inbound deltas in the inbound queue prior to local deltas in the local queue to constrain average latency in delta processing. Alternatively, the opposite order in which local deltas are processed with priority may avoid long latency in the event of a long inbound delta queue upon reconnection of the peer unit to the network.
In step 1124, the dynamics manager 804A determines the order of delta processing, and then involves the subroutine 1110 of
In this embodiment, the ordering of execution of deltas in the event of a collision depends on the endpoint that originated the delta, although other implementations can alternatively use other priority schemes or default ordering to resolve collisions. As implemented, each endpoint that is a member of a telespace is assigned (e.g., by the first endpoint member of the telespace) a unique endpoint number, corresponding, for example, to the order in which the endpoint originally joined the telespace. Thus, if peer unit 314A was the first in the telespace, it may be numbered EP1, and successive new members receive numbers according to a hierarchical numbering scheme, for example, reflecting the sponsoring endpoint that invited the new member into the telespace. Thus, if endpoint EP1 invited two new members into the telespace, they may be numbered EP1.1 and EP1.2. Then, if endpoint EP1.1 invited three new members to join, the new members may be numbered EP1.1.1, EP1.1.2, and EP1.1.3. The digits from left to right in the endpoint numbers can be referred to as first order, second order, etc. Thus, in EP1.1.3, the number “one” on the left is the first order, the “one” in the middle digit position is the second order and the “three” is the third order.
The foregoing numbering scheme has special qualities, particularly suited for assigning unique designations to endpoints for use in a peer-to-peer, fully distributed collaboration system, where members can join or discontinue membership at any time or connect or disconnect temporarily from the telespace at any time. As noted, this scheme assigns a unique designation to each endpoint of each of the telespace members. Each designation of a member indicates the order in which the member joined the telespace, and, for each member invited to join (or otherwise sponsored) by another telespace member (the “inviting member”), it also indicates the inviting member. As more and more members are invited to join by those new members, the designations add additional orders of digits to indicate the family tree or chain of inviting members. Accordingly, the designations can be thought of as indicating the ancestry of each member added to a telespace after the founding members of a telespace. Thus, where the designations comprise a number of orders, the first order designates a founding member of the telespace, and each successive order after the first order (e.g., second, third, etc, order) designates a member invited to join the telespace by an inviting member specified by the preceding order digit. The inviting member may be a founding member or another member of the telespace of intermediate seniority and with telespace membership sponsoring privileges. In one embodiment, each inviting member is responsible for assigning telespace numbers to members it sponsors.
With this numbering scheme, in step 1126 of
Thus, this scheme of dependency conflict resolution assures that any endpoint number having, for example, a “one” in the first order will be processed ahead of an endpoint number with a “two”, “three” or other higher value in the first order. Where deltas have the same number value in the first order, then the second order digits are compared, etc., until an order is specified, and then that order of delta execution is realized by performing a “do”, and, where necessary, “undo” and “redo”, operations on the deltas.
For example, the delta log 1002 may for illustrative purposes contain deltas and related information in the form
In this example, delta A has a sequence number 1 from endpoint 1.2, which is dependent on sequence number 0 from endpoint 1. Both deltas B and C are dependent on the same delta from the same endpoint, and that delta is delta A. Delta B was generated at endpoint 1.2, and delta C was generated at endpoint 1. According to the present embodiment, all three deltas are dependent deltas, and are processed in the following order: First delta A is processed because it was the first of these deltas to be received, and accordingly, the dynamics manager causes the appropriate engine to perform the data change specified by delta A. Then, delta B (as the next received delta) is processed by causing the appropriate engine to perform the data change specified by delta B. Then the dynamics manager processes delta C and notes that there is a dependency collision with delta B and that delta C has a lower endpoint number (i.e., EP1) than delta B (i.e, EP1.2). Consequently, delta B is “undone”, or more specifically, the data change that was made to the stored data is unmade so as to restore or roll-back the data to its state prior to executing that data change. Then, delta C is processed and the data change specified by delta C effected. Finally, delta B is reprocessed and its data change remade. Accordingly, the order of execution of the requested data changes in this example, after roll-back and remaking, is delta A, then delta C, then delta B.
The foregoing example relates to endpoint numbers having first and second order numbers only, though the principles illustrated apply equally well to other endpoint numbers having additional, higher orders, where collision resolution requires comparison of these higher orders, as described above. Accordingly, priority upon dependency collision is assigned in the order of membership sponsorship and then chronology based on membership date as described above.
Alternatively, priority upon dependency collision can be based on endpoint numbering assigned according to characteristics of the endpoint (e.g., job title, work group or department), nature of change requested (e.g., delete paragraph or add chapter), location of change in data (e.g., change document title or change document summary), or other considerations. To illustrate the principle behind each of these bases for awarding priority, deltas from endpoints having, for example, certain characteristics can be processed in cases of collision ahead of those having other characteristics, such as the deltas from a product design team prior to the deltas from a marketing team or vice versa.
Moreover, since roll-back and remaking of data changes (sometimes called UNDOs and REDOs) are expensive in terms of overhead, it may be desirable to provide in some implementations still other forms of prioritization of delta processing. Implementations may provide alternative priority strategies, for example, depending on activity. Some activities can have one type of conflict resolution or basis for awarding priority, while others can have other types, depending, for example, on the significance of data consistency for the particular activity. Real-time activities like chat rooms may avoid roll-back and remaking of changes in order to lower response time, particularly since order of execution may be less important.
Other implementations may execute deltas in delta clusters where they are related in some way, e.g., to changes to a certain part of a document being edited or to a certain type of change, so that they all may be made at about the same time. Furthermore, excessively-delayed deltas that have sequence numbers earlier by a predetermined value than the highest sequence number for the endpoint stored in the vector 1006 can be discarded in some implementations to avoid thrashing due to excessive redo's.
The foregoing method of processing deltas so as to optimize data consistency is particularly advantageous in implementations, such as that discussed herein, wherein A) peer units may be temporarily disconnected from the network and thus may not receive deltas until after a long delay during which local deltas may be processed out of order; B) endpoints may dynamically join or fall off membership lists; and C) overhead associated with data consistency is minimized and distributed among the peer units.
Yet other endpoint numbering schemes can be employed advantageously in particular embodiments of the invention. For example, a scheme could be implemented that uses a pseudo-random number generator to assign unique telespace member identifiers. Uniqueness can be guaranteed if the random number generator generates adequately random numbers and the seed is fairly unique. For example, using a network adapter's MAC address as a seed for the generator and creating a 32 byte value can guarantee uniqueness not only within the telespace but also within the ABC system.
The ABC system can be multitasking and multithreading in that each tool-engine combination forming an activity can correspond to a separate dynamics manager, and multiple dynamics managers can be provided to accommodate multiple simultaneously operating activities. While a one-to-one relationship between telespaces and dynamics managers may introduce the least latency in handling deltas, thread starvation could give rise to an error or fault condition. On the other hand, single dynamics managers supporting many telespaces could give rise to congestion problems. The optimal numbers of each component will depend on the hardware application, number of participating peer units, and traffic conditions.
The ABC system can comprise plural computer systems (CS), each with one or more telespaces (TS), each instantiating one or more activity components, each activity component having one or more tool (T), engine (E) pairs, each of whose operation is coordinated by a dynamics manager (DM). Thus, the ABC system can have multiple telespaces with one or more dynamics managers. Accordingly, the ABC system can be formed, for example, as follows:
CS 1 =DM 1 +A 1, where A 1 =T 1 E 1
CS 2 =DM 2 +A 2, where A 2 =T 2 +E 2
. . . .
. . . .
CS n =DM n +A n, where A n =T n +E n
where all the activities A, A2 and An reside in one or more telespaces, and “n” is a positive integer. Thus, by way of illustration, the following are examples of various telespace instantiations:
TS⊂A 1, or
TS⊂A 1 +A 2 or
TS⊂A 1 +A 2 . . . A m.
TS 1 ⊂A 1
TS 2 ⊂A 2.
TS m ⊂A m
where the symbol “⊂” means “a set comprising”, and “m” is a positive integer. Accordingly, each telespace can instantiate of one or more activities, and the framework can have a single telespace; alternatively, each telespace can instantiate of one or more activities, and the framework can have multiple telespaces.
A software implementation of components of the above-described embodiment may comprise computer instructions and routines either fixed on a tangible medium, such as a computer-readable media, e.g. the diskette 142, CD-ROM 147, ROM 115, or fixed disk 152 of
Although an exemplary embodiment of the invention has been disclosed, it will be apparent to those skilled in the art that various changes and modifications can be made which will achieve some of the advantages of the invention without departing from the spirit and scope of the invention. It will be obvious to those skilled in the art that other components performing the same functions may be suitably substituted. Further, the methods of the invention may be achieved in either all software implementations, using the appropriate processor instructions, or in hybrid implementations that utilize a combination of hardware logic and software logic to achieve the same results. Further, aspects such as the size of memory, the specific configuration of logic and/or instructions utilized to achieve a particular function, as well as other modifications to the inventive concept are intended to be covered by the appended claims.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US5008853||Dec 2, 1987||Apr 16, 1991||Xerox Corporation||Representation of collaborative multi-user activities relative to shared structured data objects in a networked workstation environment|
|US5170480||Sep 25, 1989||Dec 8, 1992||International Business Machines Corporation||Concurrently applying redo records to backup database in a log sequence using single queue server per queue at a time|
|US5392400||Jul 2, 1992||Feb 21, 1995||International Business Machines Corporation||Collaborative computing system using pseudo server process to allow input from different server processes individually and sequence number map for maintaining received data sequence|
|US5434994||May 23, 1994||Jul 18, 1995||International Business Machines Corporation||System and method for maintaining replicated data coherency in a data processing system|
|US5446842||Feb 26, 1993||Aug 29, 1995||Taligent, Inc.||Object-oriented collaboration system|
|US5538255||Feb 18, 1992||Jul 23, 1996||Barker; Bruce J.||Remote controlled multiplayer video game|
|US5555404||May 26, 1995||Sep 10, 1996||Telenor As||Continuously available database server having multiple groups of nodes with minimum intersecting sets of database fragment replicas|
|US5667440 *||Dec 20, 1994||Sep 16, 1997||Pioneer Electronic Corporation||Bidirectional communication system|
|US5737601||Jan 16, 1997||Apr 7, 1998||Oracle Corporation||Method and apparatus for peer-to-peer data replication including handling exceptional occurrences|
|US5765171||Dec 29, 1995||Jun 9, 1998||Lucent Technologies Inc.||Maintaining consistency of database replicas|
|US5765717||Apr 9, 1997||Jun 16, 1998||Gottselig; John C.||Wet hygienic towel dispenser|
|US5787262||Jun 26, 1996||Jul 28, 1998||Microsoft Corporation||System and method for distributed conflict resolution between data objects replicated across a computer network|
|US5794252||Dec 16, 1996||Aug 11, 1998||Tandem Computers, Inc.||Remote duplicate database facility featuring safe master audit trail (safeMAT) checkpointing|
|US5802322||Jun 27, 1996||Sep 1, 1998||International Business Machines Corp.||Method and apparatus for the serialization of updates in a data conferencing network|
|US5806075||Jan 14, 1997||Sep 8, 1998||Oracle Corporation||Method and apparatus for peer-to-peer data replication|
|US5823879 *||Dec 3, 1996||Oct 20, 1998||Sheldon F. Goldberg||Network gaming system|
|US5828843 *||Mar 20, 1997||Oct 27, 1998||Mpath Interactive, Inc.||Object-oriented method for matching clients together with servers according to attributes included in join request|
|US5892949||Aug 30, 1996||Apr 6, 1999||Schlumberger Technologies, Inc.||ATE test programming architecture|
|US5898834 *||Aug 16, 1996||Apr 27, 1999||Starwave Corporation||System, method, and medium for control of objects in a multi-platform environment|
|US5899810||Jan 24, 1997||May 4, 1999||Kaon Interactive Corporation||Distributed game architecture to overcome system latency|
|US5918019 *||Jul 29, 1996||Jun 29, 1999||Cisco Technology, Inc.||Virtual dial-up protocol for network communication|
|US6012096 *||Apr 23, 1998||Jan 4, 2000||Microsoft Corporation||Method and system for peer-to-peer network latency measurement|
|US6016478 *||Aug 13, 1996||Jan 18, 2000||Starfish Software, Inc.||Scheduling system with methods for peer-to-peer scheduling of remote users|
|US6030288 *||Sep 2, 1997||Feb 29, 2000||Quixotic Solutions Inc.||Apparatus and process for verifying honest gaming transactions over a communications network|
|US6104871 *||Apr 30, 1996||Aug 15, 2000||International Business Machines Corporation||Utilizing batch requests to present membership changes to process groups|
|US6112181 *||Nov 6, 1997||Aug 29, 2000||Intertrust Technologies Corporation||Systems and methods for matching, selecting, narrowcasting, and/or classifying based on rights management and/or other information|
|US6119147||Jul 28, 1998||Sep 12, 2000||Fuji Xerox Co., Ltd.||Method and system for computer-mediated, multi-modal, asynchronous meetings in a virtual space|
|US6192395 *||Dec 17, 1999||Feb 20, 2001||Multitude, Inc.||System and method for visually identifying speaking participants in a multi-participant networked event|
|US6199116||May 24, 1996||Mar 6, 2001||Microsoft Corporation||Method and system for managing data while sharing application programs|
|US6301589||Dec 2, 1998||Oct 9, 2001||Hitachi, Ltd.||Replication method|
|US6321252||Jul 17, 1998||Nov 20, 2001||International Business Machines Corporation||System and method for data streaming and synchronization in multimedia groupware applications|
|US6353596 *||Apr 12, 1996||Mar 5, 2002||Lucent Technologies Inc.||System and method for multipoint-to-multipoint multicasting|
|US6449622||Apr 15, 1999||Sep 10, 2002||Starfish Software, Inc.||System and methods for synchronizing datasets when dataset changes may be received out of order|
|US6463078 *||Jul 22, 1998||Oct 8, 2002||Microsoft Corporation||Method for switching protocols transparently in multi-user applications|
|US6480506 *||Apr 15, 1999||Nov 12, 2002||Sharewave Inc||Co-location negotiation scheme for wireless computer networks|
|US6564261 *||May 9, 2000||May 13, 2003||Telefonaktiebolaget Lm Ericsson (Publ)||Distributed system to intelligently establish sessions between anonymous users over various networks|
|US6571278||Oct 22, 1999||May 27, 2003||International Business Machines Corporation||Computer data sharing system and method for maintaining replica consistency|
|US6573926 *||Jun 21, 1999||Jun 3, 2003||Nec Corporation||Method for distributing documents in conference system|
|US6751669 *||Mar 24, 1997||Jun 15, 2004||Avaya Technology Corp.||Multimedia multiparty communication system and method|
|US20020073163 *||Dec 12, 2000||Jun 13, 2002||Fuji Xerox Co., Ltd.||Systems and methods for supporting document centered discussion across heterogeneous devices and displays|
|US20020076025 *||Dec 18, 2000||Jun 20, 2002||Nortel Networks Limited And Bell Canada||Method and system for automatic handling of invitations to join communications sessions in a virtual team environment|
|US20030069874 *||May 5, 2000||Apr 10, 2003||Eyal Hertzog||Method and system to automate the updating of personal information within a personal information management application and to synchronize such updated personal information management applications|
|US20030158900 *||Feb 5, 2002||Aug 21, 2003||Santos Richard A.||Method of and apparatus for teleconferencing|
|EP0794646A2||Mar 5, 1997||Sep 10, 1997||International Business Machines Corporation||Data management system and method for replicated data|
|EP0891064A2||Jun 9, 1998||Jan 13, 1999||Hewlett-Packard Company||Method of configuring network device|
|GB2303946A||Title not available|
|JPH0552972A||Title not available|
|JPH0991185A||Title not available|
|JPH03256146A||Title not available|
|JPH08509825A||Title not available|
|WO1997036406A1||Mar 24, 1997||Oct 2, 1997||Nokia Telecommunications Oy||Method of assigning addresses in nodes of a telecommunication network|
|WO1998038594A1||Feb 11, 1998||Sep 3, 1998||David Hingston||A system and method for distributed collaborative drawing|
|1||*||"How to Disappear Completely: A Survey of Private Peer-to-Peer Networks"-Michael Rogers and Saleem Bhatti-Walker, University College London, 1998 http://docs.google.com/viewer?a=v&q=cache:3eesq2zStZwJ:citeseerx.ist.psu.edu/viewdoc.|
|2||*||"Peer to Peer Information Systems" Gerhard Weikum http://www.mpi-inf.mpg.de/departments/d5/teaching/ws03-04/p2p-data/10-21-intro.pdf.|
|3||*||"How to Disappear Completely: A Survey of Private Peer-to-Peer Networks"—Michael Rogers and Saleem Bhatti-Walker, University College London, 1998 http://docs.google.com/viewer?a=v&q=cache:3eesq2zStZwJ:citeseerx.ist.psu.edu/viewdoc.|
|4||*||"Peer to Peer Information Systems" Gerhard Weikum http://www.mpi-inf.mpg.de/departments/d5/teaching/ws03—04/p2p-data/10-21-intro.pdf.|
|5||Begole, James "Bo", "Flexible Collaboration Transparency", Department of Comuter Science, Virginia Tech, May 12, 1997, pp. 1-30, Internet address http://simon.cs.vt.edu/~begolej/Papers/Prelim/FlexibleCollabTrans.pdf0.|
|6||Begole, James "Bo", "Flexible Collaboration Transparency", Department of Comuter Science, Virginia Tech, May 12, 1997, pp. 1-30, Internet address http://simon.cs.vt.edu/˜begolej/Papers/Prelim/FlexibleCollabTrans.pdf0.|
|7||Berglund, et al.; Amaze: A Multiplayer Computer Game; IEEE Software; May 2, 1985; p. 30-39; New York, New York.|
|8||EP Search Report, dated Sep. 24, 2009; pp. 7.|
|9||International Business Machines, "Automatic Assignment of Unique Client Identifiers", Kenneth Mason Publications, Hampshire, GB, vol. 407, No. 70, Mar. 1998.|
|10||Japanese Office Action from Application No. 2001-511554, dated Jun. 8, 2010.|
|11||Koch, Michael, "Design Issues and Model for a Distributed Multi-User Editor", Computer Supported Cooperative Work, 1995, 359-378, vol. 3, Kluwer Academic Publishers, Boston, MA.|
|12||Sun, Chengzheng, et al., "Achieving Convergence, Causality-Preservation, and Intention-Preservation in Real-time Cooperative Editing Systems", ACM Transactions on Computer-Human Interaction, Mar. 1998, pp. 2-42, vol. 5, No. 1.|
|13||Yang, Y., et al., "A Web-Based Synchronous Cooperative Editor for Software Development", ICSE-98 Workshop on Software Engineering over Internet address www3.cm.deakin.edu.au/~yun/ICSE-W98/editor/index.htm, 1998.|
|14||Yang, Y., et al., "A Web-Based Synchronous Cooperative Editor for Software Development", ICSE-98 Workshop on Software Engineering over Internet address www3.cm.deakin.edu.au/˜yun/ICSE-W98/editor/index.htm, 1998.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US8858322||Mar 13, 2013||Oct 14, 2014||Gamesys Ltd.||Systems and methods for referral benefits|
|US9055043 *||Oct 23, 2012||Jun 9, 2015||Qualcomm Incorporated||Adaptive media sharing|
|US20130124691 *||Oct 23, 2012||May 16, 2013||Qualcomm Incorporated||Adaptive media sharing|
|US20140023061 *||Sep 20, 2013||Jan 23, 2014||Canon Kabushiki Kaisha||Communication apparatus and communication method|
|U.S. Classification||709/205, 370/352, 709/224, 709/223, 370/452, 709/227, 709/203, 370/401, 370/355|
|International Classification||G06Q10/10, G06F12/00, G06F17/30, H04L29/06, G06F13/00, H04L12/58, H04L29/08|
|Cooperative Classification||Y10S707/99952, H04L51/043, H04L12/5815, H04L29/06, G06F17/30168, G06Q10/10, H04L69/329, H04L67/1095|
|European Classification||H04L29/08N9R, H04L29/06, G06Q10/10, G06F17/30F7S1|
|Feb 29, 2008||AS||Assignment|
Owner name: MICROSOFT CORPORATION, WASHINGTON
Free format text: MERGER;ASSIGNOR:GROOVE NETWORKS, INC.;REEL/FRAME:020582/0111
Effective date: 20070420
Owner name: MICROSOFT CORPORATION,WASHINGTON
Free format text: MERGER;ASSIGNOR:GROOVE NETWORKS, INC.;REEL/FRAME:020582/0111
Effective date: 20070420
|Dec 9, 2014||AS||Assignment|
Owner name: MICROSOFT TECHNOLOGY LICENSING, LLC, WASHINGTON
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:MICROSOFT CORPORATION;REEL/FRAME:034541/0477
Effective date: 20141014
|Feb 25, 2015||FPAY||Fee payment|
Year of fee payment: 4